Study on the Critical Growth Period and Yield Dynamic Forecast of Cotton in Shihezi
Cotton is one of the main economic crops in Xinjiang,and the dynamic forecast of cotton yield is of great significance for production safety.Using the daily meteorological data and cotton yield data of four ground meteorological reference stations in Shihezi area of Xinjiang from 1968 to 2020,and based on the integral regression method,the main meteorological factors and key periods of temperature,light and water affecting cotton production were analyzed with ten-day as the time scale during the whole growth period of cotton.The dynamic forecast models of cotton meteorological yield in Shihezi area in mid-July,mid-August and mid-September were established.The results showed that temperature had the greatest impact on cotton yield in Shihezi,Xinjiang.The seedling stage,flowering stage and initial stage of boll opening were the key temperature periods for cotton growth.The positive effects were significant at seedling stage and boll opening stage,while the negative effects were significant at flowering stage.The flowering stage was the key period of light for cotton growth,which had a positive effect on cotton yield.Shihezi area was an irrigated agricultural area,although the natural precipitation had a positive effect,the impact of precipitation on cotton yield was small.The dynamic forecast model established by the integral regression method was used to test the cotton yield in Shihezi region from 2018 to 2020,the results showed that the average accuracy rates in mid-July,mid-August and mid-September were 85.1％,91.4％and 94.3％,respectively.The accuracy of the cotton meteorological yield dynamic forecast model based on integral regression method was higher as it approached the mature stage.It is feasible to use the principle of integral regression to dynamically forecast cotton yield.And it can be applied to cotton yield forecasting business,providing reference for local yield forecast.
integral regression methodcottoncritical periodyielddynamic forecast
万方数据
维普
